Discrete Time State-Space Aeroservoelastic Modeling using FUN3D

CFD-based reduced order modeling (ROM) has been an active research area, as they can be used directly with common linear flutter analysis tools. Among them, linearized reduced-order modeling approaches rely on linearization of the nonlinear unsteady aerodynamic flow equations, assuming that the amplitude of the unsteady motion is limited to small perturbations about the nonlinear steady-state flow condition. Various approaches of linearized ROMs, such as Auto-Regressive-Moving-Average (ARMA), first order Volterra Kernel, Impulse Response method, etc., can be broadly found in literature. However, few of them are geared towards the controller design oriented plant modeling, i.e., to obtain a plant model with control surface actuator modeling and gust excitations, and various types of sensor definitions including sectional/component load monitoring capability. In light of this, ZONA proposes to develop a discrete time state-space aeroservoelastic modeling technique with component load monitoring using NASA developed high fidelity Navier-Stokes flow solver, FUN3D. The subspace realization algorithm will be utilized to identify the individual aerodynamic systems, i.e., due to the structural deformations (modal coordinates), control surface deflections and discrete gust, respectively. The dataset needed for the aerodynamic system identifications are obtained by a wrapper program, called OVERFUN, driving the underlying FUN3D solver. OVERFUN’s trim or static aeroelastic analysis solution will provide an initial background solution accounting for static aeroelastic effects. This unique initial flow solution sets the proposed efforts apart from other research work where an initial flow around rigid configurations is normally assumed. Once all three sub aerodynamic system are identified, they are coupled with the structural equation of motion represented in modal space and actuator models to yield the conventional state-space forms of aeroelastic model and plant model. More »